Memory MCP Server
A Model Context Protocol server that provides knowledge graph management capabilities. This server enables LLMs to create, read, update, and delete entities and relations in a persistent knowledge graph, helping AI assistants maintain memory across conversations. This is a Swift implementation of the official TypeScript Memory MCP Server.
Note: This Swift version requires macOS 14.0 or later. For broader platform support (including Windows, Linux, and macOS 11.0+), check out the Go language implementation: [memory-mcp-server-go](https://github.com/okooo5km/memory-mcp-server-go).
Swift Platform
License
Features
- Knowledge Graph Storage: Maintain a persistent graph of entities and their relationships
- Entity Management: Create, retrieve, update, and delete entities with custom types
- Relation Tracking: Define and manage relationships between entities in active voice
- Observation System: Add and remove observations about entities over time
- Powerful Search: Find relevant nodes by name, type, or observation content
- Persistent Storage: Data persists between sessions in a simple JSON format
Available Tools
create_entities- Create multiple new entities in the knowledge graph
create_relations- Create multiple new relations between entities
add_observations- Add new observations to existing entities
delete_entities- Delete multiple entities and their associated relations
delete_observations- Delete specific observations from entities
delete_relations- Delete multiple relations from the knowledge graph
read_graph- Read the entire knowledge graph
search_nodes- Search for nodes in the knowledge graph based on a query
open_nodes- Open specific nodes in the knowledge graph by their names
Installation
Option 1: One-Line Installation (curl)
The easiest way to install is with the one-line installer, which automatically downloads the latest version and installs it to
~/.local/bin in your home directory:The installer will:
- Create
~/.local/binif it doesn't exist
- Add this directory to your PATH (in .zshrc or .bashrc)
- Download and install the latest version
- Make the binary executable
Option 2: Build from Source
- Clone the repository:
- Build the project:
- Install the binary:Make sure
~/.local/binis in your PATH by adding to your shell configuration file:
Command Line Arguments
The server supports the following command line arguments:
-h, --help: Display help information about the server, its usage, and available options
-v, --version: Display the version number of the memory-mcp-server
Example usage:
Configuration
Environment Variables
The server supports the following environment variables:
MEMORY_FILE_PATH: Custom path for storing the knowledge graph (optional)
You can check the configured file path in the server logs when starting the server.
Configure for Claude.app
Add to your Claude settings:
Configure for Cursor
Add the following configuration to your Cursor editor's Settings - mcp.json:
cursor-settings
Configure for Chatwise
Add the memory MCP server to your Chatwise Settings - Tools.
chatwise-settings-tools
Example System Prompt
You can use the following system prompt to help Claude utilize the memory-mcp-server effectively:
Development Requirements
- Swift 6.0 or later
- macOS 14.0 or later
- MCP Swift SDK 0.2.0 or later
Knowledge Graph Structure
The Memory MCP Server uses a simple graph structure to store knowledge:
- Entities: Nodes in the graph with a name, type, and list of observations
- Relations: Edges between entities with a relation type in active voice
- Observations: Facts or details associated with entities
The knowledge graph is persisted to disk as a line-delimited JSON file, where each line is either an entity or relation object.
Usage Examples
Creating Entities
Creating Relations
Adding Observations
Searching Nodes
Opening Specific Nodes
Use Cases
- Long-term Memory for AI Assistants: Enable AI assistants to remember user preferences, past interactions, and important facts
- Knowledge Management: Organize information about people, places, events, and concepts
- Relationship Tracking: Maintain networks of relationships between entities
- Context Persistence: Preserve important context across multiple sessions
- Journal and Daily Logs: Maintain a structured record of events, activities, and reflections over time, making it easy to retrieve and relate past experiences chronologically
Version History
See GitHub Releases for version history and changelog.
Support the Project
If you find Memory MCP Server helpful, please consider supporting its development:
- Star the project on GitHub
- Report bugs or suggest features
- Support via:
License
memory-mcp-server is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License.
About
A Swift implementation of a knowledge graph memory server for Model Context Protocol (MCP), enabling persistent memory capabilities for large language models. This project is based on the official TypeScript implementation but rewritten in Swift using the MCP Swift SDK.
Memory MCP Server
A Model Context Protocol server that provides knowledge graph management capabilities. This server enables LLMs to create, read, update, and delete entities and relations in a persistent knowledge graph, helping AI assistants maintain memory across conversations. This is a Swift implementation of the official TypeScript Memory MCP Server.
Note: This Swift version requires macOS 14.0 or later. For broader platform support (including Windows, Linux, and macOS 11.0+), check out the Go language implementation: [memory-mcp-server-go](https://github.com/okooo5km/memory-mcp-server-go).
Swift Platform
License
Features
- Knowledge Graph Storage: Maintain a persistent graph of entities and their relationships
- Entity Management: Create, retrieve, update, and delete entities with custom types
- Relation Tracking: Define and manage relationships between entities in active voice
- Observation System: Add and remove observations about entities over time
- Powerful Search: Find relevant nodes by name, type, or observation content
- Persistent Storage: Data persists between sessions in a simple JSON format
Available Tools
create_entities- Create multiple new entities in the knowledge graph
create_relations- Create multiple new relations between entities
add_observations- Add new observations to existing entities
delete_entities- Delete multiple entities and their associated relations
delete_observations- Delete specific observations from entities
delete_relations- Delete multiple relations from the knowledge graph
read_graph- Read the entire knowledge graph
search_nodes- Search for nodes in the knowledge graph based on a query
open_nodes- Open specific nodes in the knowledge graph by their names
Installation
Option 1: One-Line Installation (curl)
The easiest way to install is with the one-line installer, which automatically downloads the latest version and installs it to
~/.local/bin in your home directory:The installer will:
- Create
~/.local/binif it doesn't exist
- Add this directory to your PATH (in .zshrc or .bashrc)
- Download and install the latest version
- Make the binary executable
Option 2: Build from Source
- Clone the repository:
- Build the project:
- Install the binary:Make sure
~/.local/binis in your PATH by adding to your shell configuration file:
Command Line Arguments
The server supports the following command line arguments:
-h, --help: Display help information about the server, its usage, and available options
-v, --version: Display the version number of the memory-mcp-server
Example usage:
Configuration
Environment Variables
The server supports the following environment variables:
MEMORY_FILE_PATH: Custom path for storing the knowledge graph (optional)
You can check the configured file path in the server logs when starting the server.
Configure for Claude.app
Add to your Claude settings:
Configure for Cursor
Add the following configuration to your Cursor editor's Settings - mcp.json:
cursor-settings
Configure for Chatwise
Add the memory MCP server to your Chatwise Settings - Tools.
chatwise-settings-tools
Example System Prompt
You can use the following system prompt to help Claude utilize the memory-mcp-server effectively:
Development Requirements
- Swift 6.0 or later
- macOS 14.0 or later
- MCP Swift SDK 0.2.0 or later
Knowledge Graph Structure
The Memory MCP Server uses a simple graph structure to store knowledge:
- Entities: Nodes in the graph with a name, type, and list of observations
- Relations: Edges between entities with a relation type in active voice
- Observations: Facts or details associated with entities
The knowledge graph is persisted to disk as a line-delimited JSON file, where each line is either an entity or relation object.
Usage Examples
Creating Entities
Creating Relations
Adding Observations
Searching Nodes
Opening Specific Nodes
Use Cases
- Long-term Memory for AI Assistants: Enable AI assistants to remember user preferences, past interactions, and important facts
- Knowledge Management: Organize information about people, places, events, and concepts
- Relationship Tracking: Maintain networks of relationships between entities
- Context Persistence: Preserve important context across multiple sessions
- Journal and Daily Logs: Maintain a structured record of events, activities, and reflections over time, making it easy to retrieve and relate past experiences chronologically
Version History
See GitHub Releases for version history and changelog.
Support the Project
If you find Memory MCP Server helpful, please consider supporting its development:
- Star the project on GitHub
- Report bugs or suggest features
- Support via:
License
memory-mcp-server is licensed under the MIT License. This means you are free to use, modify, and distribute the software, subject to the terms and conditions of the MIT License.
About
A Swift implementation of a knowledge graph memory server for Model Context Protocol (MCP), enabling persistent memory capabilities for large language models. This project is based on the official TypeScript implementation but rewritten in Swift using the MCP Swift SDK.